Our 1130 revive project. Machine from cosecans.ch

[Alex D]

Good day all,

besides the CHI-2130 we also aquired a 1130. I just thought to make a new thread for this thing.

The Machine saw power the first time. The Disk spun up but we quickly disabled it.

The printer went all to the left and continued trying to go left.

Some CPU status lights lighted normal.

The status lights on the keyboard all were on and for my liking too bright.

All voltages were okay, only +12 was more like +17V

The Machine did not react to any keypress for stop start reset, and after a short time there was minor smoke from one of the CPU cards. Have not looked into it.

Regards

Alex

[Antoine Hebert]

Hi Alex,
I am starting to have some experience with the 1130 i am restoring since june this year. I will be happy to share my minimal knowledge about this system.
We had an issue with the line filter that blew up badly.
We are no too far from each other grenoble <-> basel.

Regards,
Antoine (aconit.org)

[Alex D]

Beneath burned electronics i am beginning to smell serious collaboration here :D

If you ever come near Basel give us a heads up!

[Alex D]

Good Day,

There seems to be a Problem with the 12V and 48V wiring or switching.

Something is pulling the 12V up to around 24V measured on the light bulbs at the keyboard. Accordingly those light much too bright.

Also the 12V measured at the CPU Cage are more at ~17.

We temporarily "resolved" that by wiring the output on the 12V PSU one stage down as per the service manual, now it is at ~11V on the CPU cage. Better than 17 for the moment!

We found some cables at the 48V PSU slightly burned from heat and detoriated at the big green resistors. We replaced them and the very short test run seemed that now the lights at the KBD light even brighter. Yikes!

Disabling the 48V PSU disables the KBD lights altogether. The CPU Status Panel is lighting normal.

So, anybody might shed some light as to where the 12 and 48V might meet in the machine? As for the schematics - the Lamps all get permanent +12 and this level is okay measured directly to the 12V PSU.

But the GND is off, and somehow this results to ca. 24V over the bulb. The Machine is correctly powering off after some seconds when the 48V PSU is unhooked.

If anybody has any idea, i am as usual delighted to hear any input :)

We will be at the machine this evening if you want me to test anything.

Regards

Alex

[John Pierce]

On Fri, Oct 13, 2023 at 12:07 AM Alex D <rou...@gmail.com> wrote:

Good Day,

There seems to be a Problem with the 12V and 48V wiring or switching.

I worked at CHI in 1985 .. As far as I remember, the CPU core logic and all the peripheral logic was 74  series TTL logic, so the +5V supplies would have been by far the most important.   Any other voltages would have been for peripheral circuitry.

[John Pierce]

...................oops, hit send too soon. 

 and I forget what I was going to say...    For sure, its all way too far in the past for me to remember any specific details.

[Alex D]

Hey John,

I am talking about our newly acquired ibm 1130. the for the chi you are talking 100% correct . I think you missed my new thread :)

Regards

Alex

iPhone. iTypos. iApologize

Am 13.10.2023 um 10:04 schrieb John Pierce <pie...@hogranch.com>:



--
You received this message because you are subscribed to the Google Groups "IBM1130" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ibm1130+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/ibm1130/CAJnkzXK2LZ1VCz071eA8v6CoXw6Vfik%2BaToiGkG%2BxizjAbkMmQ%40mail.gmail.com.

[John Pierce]

On Fri, Oct 13, 2023 at 1:47 AM Alex D <rou...@gmail.com> wrote:

Hey John,

I am talking about our newly acquired ibm 1130. the for the chi you are talking 100% correct . I think you missed my new thread :)

for sure, the 1130 was early RTL..  I don't even remember what  the voltages were.

ps.google.com/d/msgid/ibm1130/2A14FBCA-9CF7-45FA-A1EB-551644834B06%40gmail.com.

[gah4]

https://en.wikipedia.org/wiki/IBM_1130

Says it is SLT, like S/360.

SLT uses +6V, +3V, and -6V power supplies.

Logic levels are TTL compatible.

(I know because I tried it.)

RTL that I remember is 3.6V. 

(Not that I used it, but I had a Popular Electronics subscription

at the time.)

On Friday, October 13, 2023 at 2:14:25 AM UTC-7 John Pierce wrote:

(snip)

for sure, the 1130 was early RTL..  I don't even remember what  the voltages were.

.

[Antoine H]

From my understanding,
If you have the midpack version (12v and 48v supply combined in one box, which i know)
12v drives keyboard, light, commands button (like program start etc.)
48v drives typewriter and disk drive solenoids, ours was not working at first but it doesn't prevent the CPU to function.
Those are ferroresonant regulated supplies
There is a third unregulated supply (don't remember the voltage) that goes to the +3 +6 -6 buck like regulators. They can be accessed by the small door under the keyboard. Each of these regulator have a circuit breaker.
The CPU uses specials open collector gates to drive (or sense ?) elements at +12 or +48v directly (instead of +3v)

Btw, we are planning to go to VCF Zurich, but we are bit late to register, so i cant confirm yet

Antoine

[Alex D]

We have the version where 12 and 48 are seperate PSUs. They are directly beneath each other.

So, the keyboard area is one where 48 and 12 meet ? We are just looking for places where this might be happening.

But as said, the common lamp +12 leads are okay in reference to the +12 ground and +12 line

The problem comes because it seems the switched negative lamp input is below 0

Maybe we measured wrong. For sure: +12 gets pulled up somewhere and the lamps see 24volts total 

iPhone. iTypos. iApologize

Am 13.10.2023 um 12:01 schrieb Antoine H <hebert.a...@gmail.com>:

From my understanding,

--
You received this message because you are subscribed to a topic in the Google Groups "IBM1130" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/ibm1130/5DwppFpnfCA/unsubscribe.
To unsubscribe from this group and all its topics, send an email to ibm1130+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/ibm1130/abe15521-57fa-4c48-b516-b005635e00e3n%40googlegroups.com.

[Carl Claunch]

SLT is a diode-transistor type of logic, not RTL. Diodes implement the logic functions and a transistor inverts the output. 

Voltage levels are 0 and +3V, not TTL levels but close enough to interface the two if you know what you are doing. The forbidden zone, switching thresholds and output range guarantees are more different than the different families of TTL, but people can get away with mixing in the right situations. 

Also a touch of SMS logic in the machine - crowbar and regulator cards in some supplies as well as the power sequencing card. SMS is discrete transistor etc on PCBs, with its own ranges of voltages.

Carl

[Antoine H]

I don't think 12 and 48 are supposed to meet anywhere, except maybe in the cpu itself. You can try measure if there is a difference between the common of the +12 and the common of the +3/+6/-6
When the cpu activates a lamp (or solenoid) it pulls the output signal to ground, eg forms check lamp  ALD_Volume_3.pdf (bitsavers.org) p79, the card is located in gate A, compartment C1, card D5 (its il all written on the machine)
There is also a 24VAC supply used as standby (in the sequence box, powered even if machine is off)
Contrary to what i said, the logic supplie takes directly line voltage at their input. As Carl said, there is also a crowbar circuit that short the output in case of an overvoltage.
The full power circuit is here :  ALD_Volume_4.pdf (bitsavers.org)

[Alex D]

Thank you! No really, thank you 😊 

There seems to be a +12 and -12 supply too.

I will look at the machine this evening later .! 

iPhone. iTypos. iApologize

Am 13.10.2023 um 18:02 schrieb Antoine H <hebert.a...@gmail.com>:

I don't think 12 and 48 are supposed to meet anywhere, except maybe in the cpu itself. You can try measure if there is a difference between the common of the +12 and the common of the +3/+6/-6

You received this message because you are subscribed to the Google Groups "IBM1130" group.

To unsubscribe from this group and stop receiving emails from it, send an email to ibm1130+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/ibm1130/c8e7517a-0eae-4b07-991d-fae725ed1919n%40googlegroups.com.

[Alex D]

As for the 24v ac - we know at least this part of the machine pretty okay for now, since this is literally the first thing to go through once you try to power it, and it is pretty similar to the chi machine :)

iPhone. iTypos. iApologize

Am 13.10.2023 um 18:09 schrieb Alex D <rou...@gmail.com>:

Thank you! No really, thank you 😊 

There seems to be a +12 and -12 supply too.

I will look at the machine this evening later .! 

[Bob Flanders]

Hi Alex, 

Where are you located?

Regards,
Bob

[Alex D]

Hey Bob,

We are located in Central Europe / Basel  :)

And there are fotos from the move of the machine from cosecans to its new home.

I am the sweaty guy with the camouflage pants 😄

Regards 

Alex 

iPhone. iTypos. iApologize

Am 13.10.2023 um 18:27 schrieb Bob Flanders <bob.fl...@gmail.com>:

Hi Alex, 

--

You received this message because you are subscribed to the Google Groups "IBM1130" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ibm1130+u...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/ibm1130/c73bbef6-7c7a-4232-a8ec-c97e92ae0e78n%40googlegroups.com.

[Ricardo Bánffy]

The thing that most annoys me here in Ireland is that Europeans seem to have been very diligent in sending old computers to recycling. I can't find even a mundane VT-230.

To view this discussion on the web visit https://groups.google.com/d/msgid/ibm1130/ECDC2B9C-A28D-4DD3-A27B-53BD53534E85%40gmail.com.

--

Ricardo Bánffy
http://about.me/rbanffy

[br...@quarterbyte.com]

That's terrific! Oscar is the one who had the DMS sources in electronic format that I was able to use to bring DMS up on our SIMH simulator! Glad to see that his collection is moving forward!

In your photographs of the move, in the second to last one, I love the row of Macintoshes on the upper level. They look like they're crowding against the railing to see what's happening down below. Are they jealous, or horrified?

Cheers

Brian

To view this discussion on the web visit https://groups.google.com/d/msgid/ibm1130/ECDC2B9C-A28D-4DD3-A27B-53BD53534E85%40gmail.com.

[Alex D]

They are waiting for command of Steve Jobs to beat up the other machines 😄. The next stations living down there may be spared, but the rest….. 

In know that Oscar is having a hard time with health issues - I initially thought he deceased but this is wrong(!). He will attend the museum on Sunday and I will take some photos. It is definitely nice to hear all the legacy about the machine and remember it.

As for the debugging issue, I mostly worked at the chi system today where there is a funky issue with output to the console.

The ibm system, we traced the 48volt line and where it meets the back of the machine it has ~18ohms to the exit of the PSU.

So there is something wrong in between. 

We checked the ground of the psu‘s because we thought this may be a … what is „star point shift“ in English? When a ground is missing and you get much higher voltages because of that….

But all grounds okay.

Regards

Alex

iPhone. iTypos. iApologize

Am 13.10.2023 um 20:43 schrieb br...@quarterbyte.com:



That's terrific! Oscar is the one who had the DMS sources in electronic format that I was able to use to bring DMS up on our SIMH simulator! Glad to see that his collection is moving forward!

In your photographs of the move, in the second to last one, I love the row of Macintoshes on the upper level. They look like they're crowding against the railing to see what's happening down below. Are they jealous, or horrified?

Cheers

Brian

On Fri, Oct 13, 2023 at 10:46 AM Alex D <rou...@gmail.com> wrote:

Hey Bob,

We are located in Central Europe / Basel  :)

And there are fotos from the move of the machine from cosecans to its new home.

I am the sweaty guy with the camouflage pants 😄

To view this discussion on the web visit https://groups.google.com/d/msgid/ibm1130/CAGBfPWLf68tmxRena_ATa2y5yF1d4PtJS6sFZC-7DODOpXeKOA%40mail.gmail.com.

[Alex D]

Guys,

Oscar was in the museum and we had a blast.
So interesting to hear everything of somebody who was there, many small stories. Did you know that his main job was not to operate the machine, but he had so much fun in tinkering with it and a very good secretary who did all of his other work 😂. So between the 3 year or so timespan when the population counts took place he had much time to play with it.

They bought the moon lander program from ibm (?) for 1000chf. Tax money 😄. But they regained this and more because once it was known they had it everybody from the other departments wanted to see it and they charged 5chf as fee.

They operated as EPL - Europe Program Library - where they shipped out and duplicated ibm software. Of course also all stuff they had not even licensed 😇. Tempting 😄

We promised him to meet again once the machine begins showing signs of life.

[Carl Claunch]

If you have -12 installed on the machine it has the Synchronous Communications Adapter on the small gate C on the right side of the machine

[Alex D]

Thank you Carl,

We def. have the SCA installed :)

currently we are more on track that somewhere 48v come to the 12v rail. I removed the Panel from the front disk drive and found the part with the Coil and the 100 Ohm resistor and the 2 small pin panels.

It seems to have gotten warm there, the legs from the smaller green resistor are corroded and a bit burnt.

We have SUCH a hard time identifying the panels, it feels like 50% of the panels are called TB3  :D . So we spent some time with a label machine and labeled everything that is 100% confirmed.

Sadly, there has been done some work and we dont know exactly what, especially in the keyboard area. Luckily, the one who did the work only had tape of a certain colour so at least we can identify it this way ;)

[Carl Claunch]

> currently we are more on track that somewhere 48v come to the 12v rail.

> I removed the Panel from the front disk drive and found the part with the Coil and the 100 Ohm resistor and the 2 small pin panels.

>It seems to have gotten warm there, the legs from the smaller green resistor are corroded and a bit burnt.

>We have SUCH a hard time identifying the panels, it feels like 50% of the panels are called TB3  :D . 

>So we spent some time with a label machine and labeled everything that is 100% confirmed.

>Sadly, there has been done some work and we dont know exactly what, especially in the keyboard area. 

>Luckily, the one who did the work only had tape of a certain colour so at least we can identify it this way ;)

IBM maintenance diagrams are often very obscure and challenging to use. 

It appears that management encouraged engineers to use the quickest method to document changes.

Instead of redrawing a diagram to ensure it is still easy to read, wires jump from page to page, with bits

of a circuit found spread over multiple sheets. 

Each team that was responsible for a part of the design would draw their own pages and they are the ones that labeled terminal blocks. Thus there are many TB-1 spread through the machine. You have to infer which it is by the area of the machine that is drawn on the page - is it the power supply or the AC/DC distribution or some other unit whose TB1 is mentioned. 

In addition, when there are multiple versions of the machine, you have to carefully interpret the small notes at the bottom of pages or know which version you have to choose the correct diagram. 

A good reference when hunting for objects inside the machine is chapter six of SY26-5977-5 , IBM 1130 Field Engineering Maintenance Manual but even this fails to show the names of the parts at the front of the disk drive.

Some components of the 1130 were designed by different engineering departments and thus documented by them, not included inside the 1130 documentation itself. 

The core storage modules (SJ-2 or SJ-4) and the internal disk drive (13SD) have their own sets of manuals and wiring diagrams, thus you are forced to jump between those and the 1130 documents, but this is also true of peripherals - 1442, 1132, 2501, 1053 etc. You will need to look at the 13SD documents to understand that circuitry. Diagram XA101 references the solenoid and resistor you are examining.

Maddeningly, that claims that the real circuit is documented back on ZB101 of the main ALDs, but it is not. The notes on ZB101 sends you to ZK111 and the circuit is not shown. However, that page does show that one lamp on the console, Disk Unlock, is wired to +48V and not to +12V. Similarly, the keyboard restore solenoid is wired to +48V.

If you suspect a connection has been made to 48V, the wires to trace out are those for the Disk Unlock lamp and for the Keyboard Restore solenoid. I would carefully look over the PKT terminal strip at the rear of the keyboard as the two voltages pass through here, but also the wiring around the Disk Unlock lamp. 

Incidentally, if you ever need to look for the resistors shown on pages such as XK111, for example the resistors for the Console Entry Switches, the resistor panel PR is underneath the console. You have to swing the machine logic gates out and crawl inside to look up to the underside to find that particular set if connections. 

The solenoid on your mystery panel is activated to release the handle for the disk drive to allow the operator to insert or remove a disk cartridge. It is driven by the 48V supply and will be active as soon as the machine powers up and does its power on reset, thus the other end of the lamp is pulled to ground by a logic gate in the rear of the disk drive. The keyboard restore solenoid is pulled to ground by logic in the 1130 logic gates but should not be active except briefly during power on reset. Look for the 48V at PKT terminals 3, 4, 8 and 20. 

Looking closer at ZK101 I can see that PKT terminal 4 is show twice on that page. It appears on the left side in the dashed line section that describes machines that had the Alpha and Numeric keys on the keyboard and would latch into either mode thus the operator did not need to hold Numeric down for very keypress delivering a numeric character. Other machines including all that I have worked on don't have this circuitry and the operator must hold down the Numeric(shift) key to get the numeric versions of a pressed key. PKT terminal 4 is also shown on the right where PKT-4 is intermixed through bail 9 on the keyboard to circuits that are powered by +12V. 

Your machine should NOT have the Alpha key on the keyboard and not have the circuitry on the left of ZK101 since the only machines that had this installed were serial numbers up to 11674 that did NOT have a SCA feature installed. Your machine has the SCA. I would still check the wiring since there might have been a field upgrade to install the SCA leaving you with a hybrid machine. 

Good luck tracking down the hidden connection.

Carl

[Alex D]

Thank you Carl, you are of great help!

We will get back to the machine on Friday ☺️

Ah, as I said in the other tread, i modified your 1130loader to fit the hex panel on our chi machine, and also wrote a small script that pipes .ipl emulator files into it, so I can load the exact same software in the emulator and the real thing. Thank you so so so much for your work!

By accident, you don’t know what the 0901 xio statement in the zrcdumc file at 4F is supposed to do, because our chi machine will only output korrekt characters when I modify it to 0900 (or 902, the problem seems to be the last bit only) after the self modification went through - and as i see it the last bits are undefined. 

Regards

Alex

iPhone. iTypos. iApologize

Am 18.10.2023 um 15:31 schrieb Carl Claunch <carlcl...@gmail.com>:

> currently we are more on track that somewhere 48v come to the 12v rail.

--
You received this message because you are subscribed to a topic in the Google Groups "IBM1130" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/ibm1130/5DwppFpnfCA/unsubscribe.
To unsubscribe from this group and all its topics, send an email to ibm1130+u...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/ibm1130/4d8ac67c-324f-4526-bb5c-1ab530196f6cn%40googlegroups.com.

[Carl Claunch]

The use of bit 15 in an XIO Write to the console printer is something specific to CHI, nothing will happen on an IBM 1130. 

The console printer logic in ALD pages XW101 to XW221 only sees two sets of signals related to XIO. It will see an XIO Write command, when the device selected is Area 1 (console) and it is T-clock step T6. It also sees when an XIO Sense for Area 1 was issued with bit 15 set (This is a sense with reset). Otherwise the logic never uses bit 15. 

The CHI machine must have different logic for its console, including some functionality when bit 15 is set during an XIO Write. Without circuit diagrams or other documentation, I can't speculate any further on what this might accomplish if anything. 

Carl

[Alex D]

Dear Carl,

Thank you -  I just wanted to make sure that there is no undocumented feature for an original 1130.

On Friday I will poke around with my scope and Logic analyzer. We have all the diagrams for the chi machine and the are much more readable than the ibm stuff 😃

Basically it is a pile of 74xxx logic which I know fairly ok because of my arcade board repairs. It is just the logic flow and I thank you for shedding some light on this too.

The schematics suppose that xio also activate a long instruction - which makes sense I guess. 

So, maybe I will get back to you with some schematics, but do not feel too pressured to help with the clone, your help with the 1130 is much more important to us 🤓

If you can, let’s shift the chi discussion to the chi thread, I would be happy to see you commenting there. As mentioned, reading the chi schematics is a piece of cake compared to the ibm machine 😄.  We have scanned some documents, the cpu and memory cage is already here:

CHI 2130 Schematic 1 of 2.pdf And 

CHI 2130 Schematic 2 of 2.pdf The video console logic is in another document, we have that too. It is a rs232c/ascii adapter. I just need to know if that is a special function or an actual error. First i will see if the set bit 15 is altering to data coming to the interface card completely or only one bit. If completely I want to poke in direction cpu instruction decoder, if only one bit i obviously need to poke direction terminal interface card I guess.  Regards Alex

iPhone. iTypos. iApologize

Am 19.10.2023 um 02:29 schrieb Carl Claunch <carlcl...@gmail.com>:

The use of bit 15 in an XIO Write to the console printer is something specific to CHI, nothing will happen on an IBM 1130. 

You received this message because you are subscribed to the Google Groups "IBM1130" group.

To unsubscribe from this group and stop receiving emails from it, send an email to ibm1130+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/ibm1130/dbf6ad54-4438-487a-91c6-1bb9c7457634n%40googlegroups.com.

[Alex D]

Hey all,

We were digging in the machine, indeed the card at „ the card is located in gate A, compartment C1, card D5“ is blown.

Our schematics, and not the one online shows a 48volt connection to said card, and there is indeed one .

Pictures speak more than words, see the wire to the said card that is wired to +48, our schematic where the 48 volt line is documented (is it plugged to the wrong spot though ? ) and our blown card .

Regards

Alex

[Alex D]

Sorry for the noise but -

Is our 48v wire connected to b4? And the schematics say b5 (??)

[Alex D]

Never mind we figured out the numbering.

The blown condensator (?) has a dead short. The rest at least is not at zero ohms 

[Alex D]

So,

what i think is happened - the short condensator (?) made a connection from +48 to ground, maybe blew the ground connection on the card and sent +48 to the base of the npn transistors.

That is mostly verified, the condensator def. connects to the 48 Volt pin with + , the other side of it def. connects to the base of the transistors.

Also we get +12 and +48 at the lamps instead of permanent +12 and switched gnd.

What i need to check is if this destroyed the GND connections on the card. 

Anybody know what the ground pin on that module would be? The condensator is fairly close to the card edge so i guess it should be connected there somewhere...

I am measuring 300Ohms to B7 from the Base and negative side of the blown condensator

And around 1 Ohm to D4 + D5. Seems not small enough for a few mm of trace. 

Going to sleep now ;)

Regards

Alex